Autonomous Intelligent Control For Path Following Of Unmanned Surface Vessels

Path following and trajectory tracking are the main problems of tracking control of unmanned surface vessels(USVs).The process of achieving control objectives of the path following of US Vs do not the requirement of time,merely satisfying position space constraints.By contrast,trajectory tracking of US Vs has strict requirement of time,and it has to satisfy constraints of USVs’position space and time.Moreover,the sailing of USVs on the open ocean only needs to move along desired path with desired velocity,and generally do not rigorously regulate the arriving time of a fixed position.The path following of USVs therefore plays a pivotal role for promoting the development of USVs.This thesis mainly focuses on the topic that is autonomou intelligent control for unmanned surface vessels’path following,underactuated ships as research plants.In the sub-system of guidance of USVs,based on Line-of-sight(LOS),three improved LOS guidance approaches are developed by considering sideslip angles that are measurable,cross-tracking errors that make a difference on LOS’s guidance performance and sideslip angles that are unknown.In the sub-system of USVs,consider problems like parameters perturbation and unknown of USV’s model,environmental disturbances,and input saturation of acturators,presenting an array of guidance-control schemes with effective performance by combining various intelligent control theories that are sliding mode control,adaptive control,designing auxiliary dynamic systems,and neural networks,etc.The main contributions of this thesis are concluded as follows:(1)For the problem of guidance laws design of path following of USVs,considering the sideslip angle that is measurable,an improved LOS(IMLOS)guidance method is presented based on existed literature such that position errors of USVs satisfy asymptotic stability.For issues of model parameter perturbation and constant disturbances of USVs,an IMLOS-based sliding mode dynamic surface adaptive control scheme is proposed through combing integral sliding mode,dynamic surface and adaptive control methods,which has realized that effective compensation for parameters perturbation and constant disturbances and avoidance the problem of calculating high-order differential operation for an intermediate virtual control law.For the troubles arising on USVs that subject to parameters perturbation and time-varying disturbances,an IMLOS-based integral sliding mode adaptive control scheme is presented by introducing integral sliding mode with power reaching laws and designing adaptive laws as well as hyperbolic tangent function used to cope with disturbances,which has realized that effective compensation for parameters perturbation and time-varyinig disturbances of USVs and solved the vibrating problem caused by the continuously switching of sliding mode surfaces.Then,the effectiveness and stability of these designed guidance-control schemes are verified by executed simulation studies for a scaled ship about I.2 meters in length and and an oil tanker about 200.6 meters in length respectively,realized effective tracking for desired velocity and desired path.(2)For the effect of the cross-tracking error on surge velocity errors,which possiblely poses a threat on the performance of IMLOS guidance methods,a velocity-varying LOS(VLOS)guidance law is developed so that position errors of USVs satisfy asymptotic stability.For USVs in the presence of lumped unknown terms that include unstructured model dynamics and errors of building model,and actuators(propellers or hydraulic propellers and rudders)saturation,a VLOS-based backstepping adaptive scheme is presented by using adaptive control and designing auxiliary dynamic systems,which realized that effective compensation for input saturation and designed adaptive laws significantly improve control performance of this method.For USVs that are running into problems of time-varying disturbances and input constraint,a VLOS and finite-time disturbances observers-based path following scheme is presented which has realized effective compensation for input saturation values that are beyond the maximum of actuators and disturbances.The effectiveness and stability of these strategies are proven by conducting simulation studies for a scaled ship about 1.2 meters in length and an oil tanker about 200.6 meters in length,achieved effective tracking for desired velocity and desired path,satisfying magnitude saturation constraints of control inputs.(3)For the problem that USVs subject to unknown sideslip angles,according to the predictor-based LOS(PLOS)and finite-time theories,a finite-time predictor-based LOS(FPLOS)guidance approach is presented,which realized that acquiring sideslip angles is faster and more stable and error system of position and sideslip errors of USVs converge with faster and better transient performance.For problems of parameters unknown totally and actuators saturation of USVs,a FPLOS-based adaptive neural network control scheme is developed.Then,the effectiveness and stability of this presented scheme is demonstrated by doing simulation experiments for a sacled ship about 1.2 meters in length and an oil tanker about 200.6 meters in length,which realized effective tracking for desired path and desired surge velocity and satisfying constraints of input saturation of USVs.The minimum learning parameters-based neural network can effectively work on learning unknown model information of USVs without any knowledge of model of USVs with merely two training weight parameters of neural networks.